Manthachitra, Vipoosit (1996) Reef fish assemblages on near-shore coral reefs: the effects of habitat structure, degradation and rehabilitation. PhD thesis, James Cook University of North Queensland.

Preview

PDF (Thesis) Download (13MB) | Preview

Abstract

The literature on the ecology of coral reef fishes is divided over the importance of habitat structure and resource availability in determining the characteristics of fish assemblages. The profile of this issue has increased, as a consequence of the increasing degradation of coral reef habitats and the need to investigate active habitat restoration as a means of reversing this process. This thesis investigates emerging generalisations about the relationships between fish communities and the characteristics of coral reef habitats on near-shore reefs, by focussing on two widely separated geographic locations (Phuket, Thailand and Orpheus Island, GBR), spatial and temporal patterns within these locations, and a range of common fish taxa (including Pomacentridae, Labridae and Chaetodontidae). The thesis employs a wide range of observational tools, including multivariate analysis to detect and describe spatial and temporal pattern in fish-habitat associations. It culminates in a series of experimental manipulations, including degradation and restoration, to test cause-effect links between fishes and different components of their habitat.

The first field programme (Chapter 3) was designed to develop an integrated transect sampling protocol for estimating fish abundance and the cover of benthic assemblages on inshore reefs. Techniques were employed to optimise sampling, not only for fish and benthic organisms separately, but also for detecting relationships between them. The performance of sampling using different transect lengths and widths was evaluated at two locations and two habitats within locations, in terms of absolute estimates, precision and efficiency. The influence of transect width on fish abundance estimates varied across localities and habitat. In contrast, transect length did not show a significant effect on estimates of abundance and shorter transects were more efficient to obtain desired levels of precision. For measuring habitat structure, two sampling techniques, Fixed Density Point (FDP) transects (in which cover was estimated using a fixed number of semi-random points) and Line Intercept (LIT) transects (in which cover was measured in terms of distance along a tape), were compared for transects of different length. The FDP method was more efficient than LIT, in terms of precision, but FDP underestimated the cover of less common habitat categories, particularly on short transects. Averaged across species, habitats and localities, a 30 m transect was considered the most efficient length for both sampling techniques. Examination of the effect of different sampling techniques and transect dimensions on the observed patterns in fish-habitat relationships indicated that only sampling with high precision (for both fish and benthos) provided consistent results. To achieve this goal, it was calculated that six replicates of a 5 x 30 m² transect were necessary for fishes and six replicates of 30 m LIT were necessary to adequately describe habitat structure. To examine relationships between the two, fish and benthic organisms were quantified using the same transect lines.

At Phuket, Thailand, the relationship of habitat structure with coral reef fish assemblages from three families; Labridae, Chaetodontidae and Pomacentridae, were investigated during 1994 and 1995 (Chapter 4). A variety of linear and parabolic relationships between living coral cover and community parameters were detected, by comparing multiple locations. Species richness was maximal at intermediate coral cover and evenness was maximal at the extremes of coral cover. Canonical Correlation Analysis identified family-wide spatial associations between fish and benthic habitat variables. In statistical terms, the Chaetodontidae were responsive (their presence/abundance depend on habitat structure variables), the Labridae were predictive (their presence/abundance indicate habitat structure variables) and the numerically dominant Pomacentridae were both responsive and predictive. Temporal variation in habitat structure, including reef degradation and unassisted recovery also influenced the composition of fish communities in predictable ways.

In the third field study (Chapter 5), the relationships between wrasse assemblage and habitat structure were investigated on fringing reefs of 3 inshore islands of the central GBR (Dunk, Orpheus and Magnetic Islands) during 1993 and 1994. Some linear and quadratic the relationships between % cover of major benthic life-forms and community parameters were detected. Living coral and algae appeared to have negative relationships with wrasse assemblages (abundance, diversity), while they were positively related to the cover of dead and/or soft coral. Canonical Correlation Analysis demonstrated significant relationships between multivariate descriptors of both fish assemblage and habitat structure. The nature of these relationships tended to be predictive for habitat structure and responsive for the fish. Temporal patterns in the structure of the wrasse assemblage were studied over a two year period at Orpheus Island. The community parameters indicated some variation at a seasonal scale, but stability over an annual scale. Canonical Discriminant Analysis indicated that, despite temporal fluctuations, spatial patterns in the structure of labrid communities persist over time.

The effects of habitat degradation on coral reef fish assemblages were experimentally investigated at two locations on Maiton Island (Phuket, Thailand - a degraded reef) and two locations on Orpheus Island (GBR, Australia - a relatively "pristine" reef) (Chapter 6). The familial structure of fish assemblages differed between these two regions, with Maiton Island co-dominated by pomacentrids and labrids, while pomacentrids were dominant on Orpheus Island. In general, fish responded negatively to living hard coral degradation, showing decreased diversity, species richness and abundance, and predictable declines in coral-associated species. The magnitude and details of the response were specific to study areas, the pre-existing condition of the habitat and taxonomic group. In contrast to hard corals, the removal of soft coral appeared to have a positive effect to most fish, apparently because domination of habitat by soft corals reduces habitat complexity.

Experimental rehabilitation of biotic habitat types was also carried out at the two geographic locations (Maiton and Orpheus Islands) to assess whether habitat rehabilitation alone was sufficient to restore fish assemblages (Chapter 7). The experiment was set up by introducing focal habitat structures (branching coral, massive corals and soft corals) to patches of degraded, largely dead coral reef. Branching Acropora induced the greatest changes in fish communities at most locations, particularly Orpheus island, where there was a consistent increase in the diversity and abundance of pomacentrids. The smaller and more site-specific effects at Maiton Island may be a consequence of the greater habitat degradation in this region, hence, restoration may be slower to act. The re-introduction of massive corals and soft corals demonstrated fewer effects, reflecting the lower physical complexity of these substrata.

The knowledge gained from this study was applied to develop a decision tool for coral reef resource management (Chapter 8). Management decisions for coral reef systems are often made on the basis of limited biological data, and status assessments are often over-simplistic, being based simply on total hard coral cover. A new assessment procedure is proposed based on a hierarchy of different levels of data availability. Four primary level indices were developed based on reef development and different components of benthic assemblage. The four indices were the Development Index (which adjusts coral cover estimates on the basis of reef development), the Condition Index (which adjusts coral cover estimates on the basis of coral assemblage condition), the Algal Index (which adjusts algal cover estimates on the basis of its potential to occupy non-living coral area), and the "Other fauna" index (which adjusts other fauna cover estimates on the basis of their potential to occupy non-living coral area). A secondary level index was also developed by integrating Development and Condition Indices, using a multi-dimensional ranking method. The application of this new procedure was carried out in three geographical regions: the east of the Gulf of Thailand, Phuket and central Great Barrier Reef. The final result is a quality rank for each study site which can be used to set up a priority list for management.

The results of this thesis support conclusions that the biotic habitat structure of coral reefs can have a major influence on the diversity and composition of reef fish communities, which parallels other marine habitats. Both natural and anthropogenic disturbances that disrupt habitats will have major secondary influences on fishes, but if necessary, at least on a small-scale, habitat-rehabilitation may reverse these impacts. The key role of habitat means that it can be developed into indices of reef condition that will also apply to fish assemblages.

Item ID:	33780
Item Type:	Thesis (PhD)
Keywords:	abundance; communities; coral reefs; degradation; ecology; fishes; Great Barrier Reef; GBR; habitats; Orpheus Island; populations; rehabilitation; restoration; structure
Date Deposited:	21 Jul 2015 04:37
FoR Codes:	06 BIOLOGICAL SCIENCES > 0602 Ecology > 060205 Marine and Estuarine Ecology (incl Marine Ichthyology) @ 70% 06 BIOLOGICAL SCIENCES > 0602 Ecology > 060202 Community Ecology (excl Invasive Species Ecology) @ 30%
SEO Codes:	96 ENVIRONMENT > 9608 Flora, Fauna and Biodiversity > 960808 Marine Flora, Fauna and Biodiversity @ 100%
Downloads:	Total: 223 Last 12 Months: 7
	More Statistics